This document discusses the frontiers of computing and predictive oncology. It provides background on the speaker and changes in computing and oncology driven by improved technology and data availability. Cancer is now defined more by underlying molecular characteristics than anatomy. Team science and open data are critical to build predictive models from large datasets and present analysis and results in a timely, human-friendly way to support treatment decisions. Key challenges include improving interoperability, validating algorithms, scaling infrastructure for large data, and reducing cognitive load in data presentation to aid decision making.